Value in the Data

I was having a conversation with a senior executive at a tier-one bank not too long ago, and the subject of big data came up. They're at a transition point, he said, where they've done the groundwork, and now they're moving to a stage where they actually want to be gaining some monetary advantage from all of the insights now. The implicit idea being that, while this analysis was all very nice, and interesting, and it kept quite a few PhDs occupied, the bills were coming due soon so could it be hurried along a little?

It's one in a number of conversations I've had lately around development and that much-loathed word, innovation. It's easy to get carried away with the new technology, but there has to be some kind of commercial benefit at the end of the day. While Goldman Sachs might be happy to call itself a technology firm, the executive said that his bank was not, and that they were a business first and foremost.

Separately, I ended up talking to a Canadian chap over the weekend who works in IT infrastructure supply for the medical sector. I started trying to relate financial services to big data and he just cut me off with a laugh, saying that if finance thinks it has big data, it has no idea. Of course, not long after this, he started cheering on Samoa in the rugby (the game on the television was England versus Australia), so take that with a pinch of salt.

The point is that, focused on technology as we are, it's easy to lose sight of the bigger picture. In GS, technology is the largest division, in JPMorgan, tens of thousands of people are employed in tech-related roles. It's big, but it's not the business. Without the actual money related parts, there wouldn't be technology divisions to speak of.

But when it comes to deriving value from data, perhaps, a greater degree of clarity is required. Data is the lifeblood of an organization, the oil in the engine. What data do you want to analyze? What questions do you want to ask it? What results do you want to find? What do you want to do with those responses? And, critically, how will that be turned into actionable intelligence to more efficiently run the business?

These are all basic questions, but often won't be used as guiding principles for data work. And the vendors love that, so when you speak about value, be precise, or you'll struggle from the word go.